| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2019-06-14 | |||
| 14:59:17 | finucannot | good catch | |
| 15:00:13 | kashyap | finucannot: Otherwise, looks good. Thanks for documenting. | |
| 15:00:46 | finucannot | done | |
| 15:00:48 | kashyap | finucannot: Speaking of docs, ... you might be horrified to know, that some of the PCIe-related command-line options in QEMU are not even documented :-( | |
| 15:00:52 | kashyap | I asked them here: | |
| 15:01:04 | openstackgerrit | Stephen Finucane proposed openstack/nova master: docs: Document how to revert, confirm a cold migration https://review.opendev.org/664396 | |
| 15:01:04 | openstackgerrit | Stephen Finucane proposed openstack/nova master: docs: Update resize doc https://review.opendev.org/665412 | |
| 15:01:08 | kashyap | https://lists.gnu.org/archive/html/qemu-devel/2019-06/msg01173.html | |
| 15:01:26 | kashyap | [Err, main e-mail here: https://lists.gnu.org/archive/html/qemu-devel/2019-06/msg01093.html] | |
| 15:01:29 | sean-k-mooney | kashyap: that is not that surprising | |
| 15:02:02 | kashyap | I'll get them documented in the official docs. I find it unacceptable that valid options are hidden like that | |
| 15:02:20 | kashyap | sean-k-mooney: They're getting better, though. From my observation. Still a long way to go. | |
| 15:02:47 | kashyap | One of the thorny "problems with documentation" is that misaligned incentives. | |
| 15:02:58 | sean-k-mooney | kashyap: ya i just am not surpirsed then the more obsucre/less used flags of complex features are not docuemtned | |
| 15:03:20 | sean-k-mooney | somethimgs its because the interfaces were experimental and they did not want to commit to it | |
| 15:03:46 | sean-k-mooney | other times it just slips through the cracks | |
| 15:04:41 | kashyap | In this case, it's officially supported, and not experimental (as you can see from the response). It was the all-too-common: "slipping through the cracks" | |
| 15:05:48 | sean-k-mooney | one of the nice things about having docs in tree is that you can require docs with each change that add a new user faceing change | |
| 15:06:06 | sean-k-mooney | or wehre you have docs generated form code | |
| 15:06:30 | efried | cdent: Will look. Been kind of avoiding because out of my realm, but I can ride on your vote with clear conscience | |
| 15:07:10 | cdent | I think in this case you can rely on the tests, not me | |
| 15:07:30 | openstackgerrit | Stephen Finucane proposed openstack/nova master: docs: Remove references to nova-consoleauth https://review.opendev.org/652965 | |
| 15:07:30 | openstackgerrit | Stephen Finucane proposed openstack/nova master: tests: Stop starting consoleauth in functional tests https://review.opendev.org/652966 | |
| 15:07:31 | openstackgerrit | Stephen Finucane proposed openstack/nova master: xvp: Start using consoleauth tokens https://review.opendev.org/652967 | |
| 15:07:31 | openstackgerrit | Stephen Finucane proposed openstack/nova master: nova-status: Remove consoleauth workaround check https://review.opendev.org/652968 | |
| 15:07:32 | openstackgerrit | Stephen Finucane proposed openstack/nova master: Remove nova-consoleauth https://review.opendev.org/652969 | |
| 15:07:32 | openstackgerrit | Stephen Finucane proposed openstack/nova master: objects: Remove ConsoleAuthToken.to_dict https://review.opendev.org/652970 | |
| 15:07:33 | openstackgerrit | Stephen Finucane proposed openstack/nova master: docs: Rework nova console diagram https://review.opendev.org/660147 | |
| 15:13:18 | efried | cdent: That wasn't so scary, found a page that explains it clearly. dansmith, would you please send https://review.opendev.org/#/c/664193/ ? | |
| 15:14:07 | openstackgerrit | Artom Lifshitz proposed openstack/nova master: Revert resize: wait for events according to hybrid plug https://review.opendev.org/644881 | |
| 15:14:07 | openstackgerrit | Artom Lifshitz proposed openstack/nova master: WIP: Include direct-physical in compute manager events check https://review.opendev.org/664431 | |
| 15:14:08 | openstackgerrit | Artom Lifshitz proposed openstack/nova master: [DNM] testing bug/1813789 revert resize events https://review.opendev.org/664442 | |
| 15:14:20 | dansmith | efried: link me to the page? because it looks scary on the surface | |
| 15:14:34 | efried | dansmith: https://docs.sqlalchemy.org/en/13/changelog/migration_09.html#new-query-options-api-load-only-option | |
| 15:16:06 | dansmith | efried: that page says that the old methods will be supported "indefinitely" | |
| 15:16:31 | cdent | they spit warnings | |
| 15:16:36 | cdent | lots and lots and lots of warnings | |
| 15:17:09 | dansmith | oh, hmm, that's annoying given the statement | |
| 15:18:28 | dansmith | cdent: your suggestion would be a lot tidier | |
| 15:18:43 | dansmith | cdent: any chance you'd just do that real quick so I and efried can approve _that_ ? | |
| 15:19:02 | dansmith | should just be a few lines of change it looks like | |
| 15:19:04 | cdent | yeah, can do | |
| 15:19:07 | cdent | one mo | |
| 15:28:04 | efried | cdent, dansmith: tbh I would prefer going the other way | |
| 15:28:17 | efried | but whatever | |
| 15:28:20 | dansmith | efried: that's not as easy | |
| 15:28:27 | cdent | i'm basically done | |
| 15:28:27 | dansmith | because we calculate column names in places | |
| 15:28:30 | cdent | yes, that | |
| 15:29:32 | efried | right, I get that, so I'm fine with it. But I actually prefer the explicitness of takashi's patch. But I'm fine either way. | |
| 15:30:14 | cdent | the new diff is way tidier | |
| 15:30:16 | dansmith | joinedload, joinedload_all, joined, etc has brought plenty of confusion in the past, so more consistency for people to copy seems like a win to me | |
| 15:30:56 | dansmith | also, cdent and I are agreeing which means (1) jump on that shit and (2) go buy some lottery tickets | |
| 15:31:08 | cdent | ✔ | |
| 15:32:32 | openstackgerrit | Chris Dent proposed openstack/nova master: Replace joinedload_all with joinedload https://review.opendev.org/664193 | |
| 15:32:49 | cdent | i didn't run any tests on that ^ | |
| 15:32:54 | cdent | will start them now | |
| 15:33:52 | dansmith | cdent: I was going to write a unit test for the helper too | |
| 15:34:20 | efried | lgtm cdent | |
| 15:34:41 | cdent | yeah, I gotta run now, so hope that's suitable, but will leave it in dan's earlier in his day hands | |
| 15:38:58 | openstackgerrit | Stephen Finucane proposed openstack/nova master: Follow-up for I6a777b4b7a5729488f939df8c40e49bd40aec3dd https://review.opendev.org/664967 | |
| 15:38:58 | openstackgerrit | Stephen Finucane proposed openstack/nova master: Remove 'MultiattachSupportNotYetAvailable' exception https://review.opendev.org/651315 | |
| 15:39:35 | finucannot | efried: I removed the additional test that I'd meant to remove in https://review.opendev.org/#/c/664967/. Fancy revisiting? | |
| 15:39:47 | efried | ... | |
| 15:40:26 | finucannot | efried: Addressed this comment https://review.opendev.org/#/c/621061/8/nova/tests/unit/compute/test_compute_api.py@4477 | |
| 15:40:35 | finucannot | in the follow-up | |
| 15:41:26 | openstackgerrit | Stephen Finucane proposed openstack/nova master: objects: Remove legacy '_from_dict' functions https://review.opendev.org/537414 | |
| 15:41:33 | efried | finucannot: +2, thanks for getting on the new=mock.Mock() bandwagon with me. | |
| 15:48:17 | dansmith | efried: you have insulted me to my core | |
| 15:48:29 | efried | payback is hell | |
| 15:49:59 | sean-k-mooney | efried: why use new=mock.mock() ? in what case is that needed? | |
| 15:50:33 | dansmith | new= prevents it from being passed as an arg | |
| 15:50:51 | efried | sean-k-mooney: When using decorator: | |
| 15:50:51 | efried | @mock.patch("...") <== causes an argument to be created to your method | |
| 15:50:51 | efried | @mock.patch("...", new=...) <== no arg | |
| 15:50:54 | efried | yeah, what dansmith said | |
| 15:51:10 | sean-k-mooney | dansmith: oh ok that makes sense if you never want to use of check it | |
| 15:51:12 | temka | efried, that's *massively* useful | |
| 15:51:22 | temka | efried, I've been using _ where possible | |
| 15:51:36 | dansmith | sean-k-mooney: yeah, but it's also a good thing to verify that the person isn't skirting some checking they should be doing as a result :) | |
| 15:51:42 | sean-k-mooney | temka: in a fucntion signiture | |
| 15:51:46 | efried | temka: You mean it's a thing you didn't know, or you're agreeing that it's useful and you use it? | |
| 15:51:55 | temka | efried, yes ;) | |
| 15:51:56 | sean-k-mooney | efried: yes | |
| 15:52:08 | temka | sean-k-mooney, yeah, you can have def blah(self, _) | |
| 15:52:23 | sean-k-mooney | temka: yes that would be an instance -1 form me | |
| 15:52:33 | temka | sean-k-mooney, you'd be alone then :) | |
| 15:52:36 | sean-k-mooney | i really think that is poor api design | |
| 15:52:45 | temka | Only in unit tests | |
| 15:53:08 | sean-k-mooney | you are technically not overloading if you cange teh name to _ | |
| 15:53:25 | sean-k-mooney | so even in unit tests you shoudl not do that | |
| 15:53:45 | temka | [artom@zoe nova]$ grep def.*_\) -R nova/tests/ | wc -l | |
| 15:53:45 | temka | 13 | |
| 15:53:49 | sean-k-mooney | i guess you are doingit for function that will be used in place of a mock? | |
| 15:53:50 | temka | Wonder if all of those are mine | |
| 15:54:30 | openstackgerrit | Dan Smith proposed openstack/nova master: Add a test for the _joinedload_all helper https://review.opendev.org/665421 | |
| 15:54:31 | dansmith | efried: ^ | |
| 15:55:50 | efried | dansmith: +2, nice. Added cdent | |
| 15:56:08 | sean-k-mooney | temka: i wonder if the new=mock.Mock works the same way when your apply the decorator to a class | |
| 15:56:17 | efried | it does | |